Plasmas from 14 patients with factor VIII inhibitors, 10 haemophiliacs and four non-haemophiliacs, were assayed by both the agarose gel and Bethesda methods. Good correlation was observed in 34 samples from 13 patients, but there was poor correlation in three samples from a single haemophilic patient. The sensitivity of the method was increased by diluting normal platelet rich plasma (PRP) with congenital factor VIII deficient plasma. With this modification, as little as 0.4 of a Bethesda unit (Bu) could be measured accurately. The agarose method is easier to perform and requires much less technician time than the Bethesda assay (Ba). Inhibitory activity can be measured even in the presence of large amounts of transfused factor VIII or factor IX concentrates. To study the effect of factor IX concentrates on the inhibitor, the method was modified by incorporating into the gels plasmas specifically deficient in either factor VII, IX or X. Our data suggest that factor VII is probably responsible for the 'bypassing' activity of factor IX concentrates.